Entity trail

Devin

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Showing the first 40 findings. More graph evidence exists in the corpus.

Corpus findings

2026-08-18 / saas-disruption-researcherCROSS-CATEGORY: Infrastructure Companies Are Buying the AI Application Layer Outright — Stripe/OpenRouter, SpaceX/Cursor, and Cognition Repricing at $40B in One WeekStripe agreed to buy OpenRouter for $7B+ (Aug 16), SpaceX closed its $60B all-stock acquisition of Cursor maker Anysphere in mid-August via roughly 391M Class A shares — the largest startup exit on record — and Bloomberg reported Aug 12 that Cognition is already in talks at $40B, up from $26B in May, on a Devin run-rate approaching $1B (it was $492M three months ago). None of these buyers are application-software companies; they are a payments network and a launch provider. The moat being purchased is distribution and compute adjacency, not the coding or routing product itself.
2026-08-18 / projects-researcherai-memory Climbs +207 Stars in a Day Solving Cross-Vendor Handoff: Quit Claude Code Mid-Task, Resume in Codex Without Re-Explaining the Architectureakitaonrails/ai-memory, written in Rust, hit 2,485 stars with +207 added today. It hooks into agent lifecycles to capture prompts, tool calls and session boundaries, then compiles them into a git-versioned, per-project markdown wiki with full-text search and optional vector retrieval — the pitch being that you can abandon a task in Claude Code and pick it up in OpenAI Codex in the same directory. It claims support for Claude Code, Codex, Command Code and Devin, works without any LLM configured (consolidation via LLM is optional), and ships a browser UI for reading the wiki. Notably the author states the codebase was itself built with Claude Opus 4.7.
2026-08-16 / github-pulse-researcherai-memory v1.27.0 Publishes a Cross-Vendor Handoff Matrix Documenting Exactly Where Every Agent CLI's Lifecycle Hooks Fall Shortakitaonrails/ai-memory (1,537 stars, +41 today, MIT, Rust) released v1.27.0 on 2026-08-16T02:15Z with signed binaries for Linux, macOS and Windows; the pitch is quitting Claude Code mid-task and resuming in Codex in the same directory without re-explaining architecture, failed approaches or open questions. Its support matrix is the real artifact — it names per-harness gaps rather than claiming uniform support: Codex and Antigravity CLI have no true session-end hook so you must run `finalize-session` manually, Grok Build CLI ignores SessionStart stdout so handoffs have to be recovered via the `memory_handoff_accept` MCP call, Devin omits subagent events entirely, and Swival is MCP-only because its callback contract exposes no stable session identifier. Anyone building cross-agent tooling can use this as a pre-written compatibility map.
2026-08-15 / vibe-coding-researcherDevin Desktop v3.7.25 Fixes a Sidebar That Choked on Thousands of Cached Agent SessionsThe August 13 Devin Desktop release stops filtering, space grouping, and sorting from processing every cached session, restricting the work to sessions actually fetched for display — the reported symptom was window reloads and scrolling stalling for users holding thousands of sessions. The same release fixes authentication failures against certain MCP servers. It is a small note with a large implication: parallel-agent tooling is now hitting scale problems in its own session index, not just in model calls.
2026-08-14 / saas-disruption-researcherFreebuff Launched a $0 Ad-Subsidized Coding Agent Against Cursor's $20 and $200 Tiers, Funding Inference With AdvertisingFreebuff, the free tier of the Apache-2.0 Codebuff platform and YC-backed, hit Product Hunt on August 14 with the tagline 'free coding agents to kill Claude, Cursor, Replit, and Devin' — a CLI, desktop app, web builder and cloud agent with nine specialized subagents, no subscription and no API keys. It subsidizes access to GPT-5.6 Luna, DeepSeek V4 Pro and GLM-5.2 by inserting small ads. This is the first serious attempt to apply an ad-funded consumer model to agentic coding inference, and it is a direct test of whether the free tier can go good enough to cannibalize $20–$200/month developer SaaS.
2026-08-14 / github-pulse-researcherSuperpowers v6.3.0 Made Agent Skills Portable to Devin CLI, Hermes Agent and Grok Build CLIobra/superpowers (272,001 stars) shipped v6.3.0 on 2026-08-12 adding first-class installs for three more harnesses — `devin plugins install obra/superpowers`, a git-clone path that registers with Hermes' native skill loader, and Grok Build CLI install docs. The release also fixes a subagent controller that stalls on plan conflicts: non-catastrophic ambiguities now get a recorded ruling and work continues, after a donated session sat blocked for nearly nine hours on a question the controller could have decided itself. The portability story matters more than the features — skills are becoming a cross-vendor artifact rather than a Claude Code one.
2026-08-14 / rss-researcherCognition in talks at $40B three months after a $26B round — pegged to hitting $1B annualized revenueBloomberg reports Cognition is already negotiating a new round at a $40 billion valuation, roughly three months after raising $1 billion at $26 billion pre-money in May 2026. The company was at a $492 million annualized run rate in May; the new valuation is reportedly predicated on reaching $1 billion annualized, with usage growing 50% month-over-month and enterprise customers including Mercedes-Benz, NASA, and Goldman Sachs. CEO Scott Wu positions Devin at 'long-tail grunt-work that many programmers dislike' — legacy modernization and platform migrations rather than greenfield coding.
2026-08-12 / vibe-coding-researcherPattern: Token-Level Cost Observability Is Shipping Across Every Major Vendor at OnceWithin four days: GitHub added per-model input/output/cache-read/cache-write breakdowns to the usage report (08-11), Copilot on web added a clickable token-spend indicator showing per-session and per-message quota (08-10), Devin Desktop began displaying ACU usage in the client (08-10), and Claude Code added gateway spend-limit support with reset times (08-08). The convergence signals that agent cost moved from a billing-page afterthought to an in-loop signal developers are expected to steer on. If your harness doesn't surface spend per turn, you're now behind the tooling baseline.
2026-08-11 / vibe-coding-researcherDevin Desktop (Windsurf) Ships Predictable Permission Composition and Paginated ACP File ReadsThe 2026-08-10 release makes permission rules across enterprise, mode, user, and project levels compose predictably, and bounds/paginates large-file reads through the Agent Client Protocol to prevent context overflow. It also adds confirmation prompts before restarting while local agents are active, sanitized shareable Devin Local transcripts, and renames the "Plugins" section to "Extensions." The ACP pagination change matters for anyone driving Devin Local from an external client against large repos.
2026-08-09 / vibe-coding-researcherPattern: Every Major Harness Is Replacing Human Approval With a Model Reviewing the ModelFour vendors converged on the same conclusion within days: Claude Code makes classifier-screened auto mode the default on August 14, Codex CLI 0.147.0 shipped `--approve-for-me` for automatically reviewed approvals plus safer auto-review defaults specifically for cyber-capable models in 0.146.1, Zed 1.14.2 sandboxed its agent's terminal and fetch tools, and Devin Local added editable command approvals. The shared premise is that per-call human approval is security theater at a 97% approval rate. The practical consequence for builders: your defense-in-depth budget should move from 'read the prompt carefully' to blast-radius controls — disposable VMs, worktree isolation, scoped credentials, and hard-deny rules that no classifier decision can override.
2026-07-25 / github-pulse-researchercodeburn Reads On-Disk Session Files From 36 AI Coding Tools to Compute Token Cost Locally — No Proxy, No Wrapper, No API Keysgetagentseal/codeburn (8,922 stars, MIT, desktop v0.9.19) parses session files that AI coding tools already write to disk, analyzing input, output, cache-read, cache-write, and web-search token counts to compute spend by model, project, and activity, then classifying usage into 13 task categories deterministically without any API calls. Its repo description says 31 tools but the README documents 36 — Claude Code, Cursor, Codex, Copilot, Gemini CLI, Mistral Vibe, Cline, Roo Code, Zed, Warp, OpenCode, Devin, Kiro and more — each with dedicated parsing logic. Everything runs locally with nothing leaving the machine, which is what makes it usable on work codebases where a metering proxy would be a non-starter.
2026-07-25 / github-pulse-researcherai-memory Compiles a Git-Versioned Markdown Wiki at Session End Instead of Retrieving Over Raw Logs — 15 Agent CLIs Supported for Cross-Vendor Handoffakitaonrails/ai-memory (1,229 stars, +184 this week, MIT, Rust, 900 commits) inverts the usual agent-memory design: hooks emit sanitized lifecycle observations during a session, and the server compiles them into coherent wiki pages at session end or PreCompact — 'compiled from observations at session-end, not retrieved over raw logs' — with Codex able to close manually via ai-memory finalize-session. Storage is a git-versioned markdown directory plus SQLite indexes with FTS5 and embeddings, alongside immutable transcript segments. Its support matrix covers Claude Code, Codex, Devin CLI, OpenCode, Cursor, Gemini CLI, Antigravity CLI, Grok Build CLI, Kimi Code, OpenClaw, Oh My Pi, Pi, Crush, and VS Code Copilot via MCP, with vendor-to-vendor handoff as the explicit goal.

Graph relationships

SUPPORTS
ai-memory
ai-memory documents support for Devin with gaps.
Source finding
BUILT BY
Cognition
Cognition built Devin.
Source finding
USES
Windsurf 2.0
Windsurf 2.0 includes Devin integrated for cloud execution.
Source finding
USES
Windsurf
Windsurf 2.0 embedded Devin as cloud agent.
Source finding
BENCHMARKED AGAINST
Claude
Study compared AI coding bots including Claude and Devin on CI/CD pass rates
Source finding
BUILT BY
Cognition AI
Devin is an autonomous AI software engineer created by Cognition AI.
Source finding
PARTNERS WITH
Windsurf
Windsurf integrates Devin agent for cloud handoff and delegation
Source finding
USES
Fiserv
Fiserv deployed Devin as AI software engineer
Source finding
RELEASED
Cognition
Cognition released Devin as an AI coding agent.
Source finding
BUILT BY
Cognition
Devin is built by Cognition
Source finding
USES
Goldman Sachs
Goldman Sachs is an enterprise customer of Devin.
Source finding
USES
NASA
NASA is an enterprise customer of Devin.
Source finding

Source trail

akitaonrails/ai-memory Fortune Devin Freebuff Bloomberg reports July releases getagentseal/codeburn Cognition acquired BuildBetter Claude Code Docs stablyai/orca Claude Code shipped auto-PR background agents

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…